Sudarsun Santhiappan, PhD

icon indicating a website link http://www.sudarsun.in icon indicating an email [email protected]

icon company IIT Madras icon location Chennai icon location Counting days to experience the singularity!

Results 3 comments of Sudarsun Santhiappan, PhD

AttributeError : HARM_CATEGORY_DEROGATORY

Yes, currently, only the Gemini family types are supported, which are the following: HARM_CATEGORY_HARASSMENT HARM_CATEGORY_HATE_SPEECH HARM_CATEGORY_SEXUALLY_EXPLICIT HARM_CATEGORY_DANGEROUS_CONTENT HARM_CATEGORY_UNSPECIFIED

wikitext-103-raw-v1.zip is not available on the amazonaws anymore

The file is available at 'https://dax-cdn.cdn.appdomain.cloud/dax-wikitext-103/1.0.1/wikitext-103.tar.gz'

VRAM spikes after "LlamaForCausalLM does not accept 'num_items_in_batch'"

On an RTX 4090 Ti, I had to set up `per_device_eval_batch_size=1` and `per_device_train_batch_size=1`; however, I could make it work when my `eval_accumulation_steps` and `gradient_accumulation_steps` were set to `8`.